Musk’s GPT Rival, Grok-3 is ‘Scary Smart’—But Why Can’t Most People Try It?

Elon Musk’s artificial intelligence venture, xAI, has just introduced its most advanced chatbot, Grok-3. It is touted as a major leap forward in AI capabilities. Built to rival OpenAI’s GPT-4 and Google’s Gemini models, Grok-3 has been trained on twice the computational power of its predecessor, promising enhanced reasoning, improved problem-solving, and more human-like responses.

However, there’s a catch. Most users won’t get to experience it. Instead of making Grok-3 widely available, xAI has restricted access exclusively to X Premium+ subscribers. This decision has sparked debates about whether limiting access could slow down AI adoption and hinder Grok-3’s potential impact.

Grok-3’s Scary Smart Capabilities

Grok-3 is described by Musk as “an order of magnitude more capable” than its predecessor, Grok-2. The model was trained using 200,000 Nvidia H100 GPUs, doubling the computational resources of Grok-2. This intensive training aimed to enhance Grok-3’s reasoning abilities, enabling it to tackle complex queries by deconstructing them into manageable tasks. Early internal benchmarks suggest that Grok-3 surpasses leading models from OpenAI, Google, and DeepSeek in areas such as mathematics, science, and coding. However, these claims await independent verification. The managing director at D.A. Davidson, Gil Luria said, “The introduction of Grok-3 puts xAI back in the race for leadership in open-source LLMs. It outperforms the current state-of-the-art models on some benchmarks, which makes xAI relevant again.”

• DeepSearch: A New Feature

Accompanying Grok-3’s launch is DeepSearch, a next-generation search engine designed to reduce the time users spend seeking information online. DeepSearch aims to provide concise, accurate responses, streamlining the research process. “Grok-3 across the board is in a league of its own,” Musk said during a livestream alongside three xAI engineers late on Monday, adding the model outperforms its predecessor, Grok-2.

Access Limitations

Despite its potential, Grok-3’s reach is currently limited. The access is granted exclusively to X Premium+ subscribers, a service priced at $40 per month or $395 annually. Additionally, xAI has introduced a SuperGrok subscription at $30 monthly or $300 annually, offering advanced features like DeepSearch and increased image generation limits. This tiered access model may deter a wider audience, particularly casual users or those unwilling to invest in premium subscriptions. Users can access the chatbot via its mobile app and Grok.com website.

• Grok-3 Subscription Fees: $40 per month (via X)
• SuperGrok Subscription Fees: $30 per month

Potential Barriers to Adoption (No Free Models) 

The decision to confine Grok-3’s access to premium subscribers raises questions about its widespread adoption. By restricting availability, xAI may inadvertently hinder the model’s exposure and user base expansion. In contrast, competitors like OpenAI and DeepSeek have embraced more inclusive approaches, offering free or open-source models that attract a diverse user demographic. The inclusion of free models promotes community engagement, feedback, and iterative improvement—elements essential for AI development.

xAI’s strategy appears to prioritize monetization and exclusivity, potentially at the expense of broader user engagement. The premium subscriptions can generate revenue, but they may also limit the model’s reach and the diversity of user interactions necessary for refining AI performance. Balancing financial objectives with accessibility could be key to Grok-3’s long-term success and relevance in the competitive AI market.

Our Take

Grok-3 represents a significant advancement in xAI’s AI capabilities, with early indications of superior performance in specialized domains. However, its restricted access model may impede widespread adoption and limit the diverse user engagement essential for continuous improvement. To establish Grok-3 as a leading AI tool, xAI might consider strategies that balance exclusivity with broader accessibility, promoting a more inclusive user community and enhancing the model’s development trajectory.

Also read:

DeepSeek AI Surpasses WhatsApp to Become Top Android App in Pakistan